This is part of the disentangling process as major network assets in Italy change hands
Swisscom’s Italian subsidiary Fastweb is to sell its 4.5% stake in FiberCop to Optics Bidco, a subsidiary of KKR. FiberCop was established in 2021 by Telecom Italia (TIM), KKR and Fastweb to develop national fibre broadband infrastructure in Italy.
KKR will acquire all shares held by Fastweb in FiberCop for a cash consideration of €438.7 million, which is in line with the pro rata price paid by KKR to TIM for its stake.
Setting up the NetCo
KKR is in the process of acquiring TIM’s fixed network in a transaction worth about €22 billion. The deal was “unconditionally” approved this week by the European Commission and has the backing of the Italian government. The government retains the right to intervene in anything that affects the so-called NetCo as it is classed as critical national infrastructure.
The sales of its stake in FiberCop to KKR’s subsidiary is subject to the completion of the NetCo transaction by KKR, which is expected to close in Q3 2024. The transaction has no effect on the wholesale agreement between Fastweb and FiberCop, according to Swisscom.
Fastweb’s future
Meanwhile, Fastweb is in the throes of the approvals process to acquire Vodafone Italy for €8 billion which it hopes will be completed next year and will transform it into fixed-mobile operator.
In the announcement today, Swisscom stated, “Fastweb remains fully committed to its mission of driving innovation and connectivity in the country through investments in key telecommunications infrastructures.
“Fastweb will therefore keep making relevant investments to increase the coverage of its proprietary, end-to-end controlled fiber network and will continue to be a key provider of wholesale services to third parties, ensuring the availability of robust and competitive offerings in the market.”
The operator group has teamed up with i2CAT to help reach its target of having 30% of its masts in Europe using Open RAN by 2030
Vodafone and the i2CAT Foundation – a technology research centre – announced that they will use the improved automation offered by RAN jointly to build a responsive, multi-vendor management system. Its purpose will be to fix faults and respond to cyber threats faster and more cheaply.
The plan is to combine the strengths of i2CAT’s R&D in new digital technologies with Vodafone’s engineering expertise at its Innovation Centre in Málaga. They will use machine learning to manage and analyse logs from multi-vendor Open RAN. These logs provide vital information, such as successful login or failed access attempts, which can be used to improve security and detect threats.
The ultimate intention is that the system will provide Vodafone with a dashboard from which it can respond to and control Open RAN events over a big geographical area. This will support the company’s aim of having 30% of its masts across Europe using Open RAN technology by 2030.
Reducing operational costs
The management system should reduce operational costs by: automating the processing and analysis of multi-vendor logs; strengthening security by being able to quickly detect and mitigate threats across different suppliers; and simplifying compliance with appropriate regulations and industry standards.
The initial focus for Vodafone and i2CAT is on the design of a Security Information and Event Management system (SIEM) to flag potential security threats like unauthorised access, denial-of-service attacks, or man-in-the-middle interceptions.
Vodafone and i2CAT will test how to differentiate diverse types of logs, classify and manage them according to the specific threat. The intention is then to integrate the SIEM with the main Open RAN components, such as the RAN Intelligence Controller (RIC), which controls and optimises the radio access network functions.
The two organisations will also run proofs of concept (PoCs) that include log management of automated rApps, which are used for applications such as steering traffic or conserving energy.
The results of this collaborative project, named Holistic ORAN Logging & Metrics Security Shield (HOLMES), will be shared with the O-RAN Alliance to enable vendors to contribute and adopt new standardised approaches to log formats.
Vodafone is committed to working with leading academic and research institutions and for example, has worked with the University of Oxford to guard against risks arising from AI developments.
Francisco Martin, Head of Open RAN at Vodafone, said, “Vodafone’s partnership with i2CAT…will enable us to automate more manual tasks associated with traditional networks to respond even faster to fluctuations in demand, manage energy consumption more effectively, launch new features quicker and keep ahead of the ever-changing threat landscape.”
Not a moment too soon
Massimo Fatato, Head of Networks at NTT Data UK&I, commented, “Vodafone’s collaboration with the i2CAT Foundation is a key indication of the direction the industry is taking…towards the evolution of telecommunications infrastructure – and not a moment too soon.
“The development of an automated Open RAN management system is a fundamental component to fulfil Open RAN’s promise of interoperability and flexibility in the access network. It will be a critical enabler as we move towards more dynamic and resilient network architectures. It will also aid Vodafone immensely in having 30% of its masts using Open RAN technology by 2030.”
He continued, “We can see this as part of a larger movement towards intelligent, software-defined networking” and telcos’ need to cut costs. “Leveraging AI in automating the management of Open RAN will not only improve operational efficiency but also accelerate the deployment of new services and features, ensuring that networks can adapt quickly to changing demands.”
Although the two have worked together for a decade, the new partnership is targeted at developing Silknet’s service portfolio
Orange has formed a new strategic partnership with Georgia’s leading fixed and mobile network provider Silknet which will cover that telco’s service capabilities in the business, consumer and ICT markets. The partnership is part of Orange’s Alliance program which launched in 2014 in French Polynesia and Portugal.
The timing is interesting given in March, the Georgian operator had a data leak impacting “up to 2000” users, although local media reports speculate that up to 33,000 were impacted with ID card copies appearing on the internet. The incident is still under investigation, but Orange Cyberdefense is certainly a part of the service portfolio Silknet may find attractive.
As part of the agreement, Silknet will gain access to Orange’s reservoir of knowledge and experience in the telecom and ICT sectors. One of the initial focuses of this alliance will be on enhancing B2B ICT services. Orange said it will provide substantial support to Silknet to “transform the telecom landscape”.
“Silknet has a 10-year history of successful cooperation with Orange, and I am happy that it has now grown into a strategic partnership,” said chairman of Silknet’s supervisory board George Ramishvili. “It marks a new era in the development of Silknet’s digital ecosystem enabling it to offer cutting-edge products and services. We are delighted that this global European conglomerate will help Silknet introduce its innovations to the Georgian and regional telecoms markets.”
“Orange is pleased to confirm and strengthen its relationship with Silknet through the Orange Alliance program,” said Orange EVP and CTIO Bruno Zerbib. “Leveraging Orange expertise, products and platforms will create value for Silknet on the Georgian market. Building scalable platforms and opening them to partners is at the core of our innovation strategy.”
Looking ahead, Silknet and Orange stated they are committed to expanding this partnership by exploring new avenues for cooperation and further integrating their operations.
An expensive market
Silknet launched 5G services last December in the capital Tbilisi. According to the Georgian Communications Commission, 100Mbps fixed broadband is 136% more expensive the the average in Euorpean countries while mobile internet with 10GB speed and 1795 minutes of talk time is 35% more expensive than the average price in Europe. The study by Strategy Analytics was based on prices for services of the two major Georgian operators, Magticom and Silknet, as these companies own more than 70% of the Georgian telecommunications market.
According to the survey results, low-volume and low-speed services in Georgia are almost similar in cost or cheaper than in Europe and Britain. However, prices for high volume and high-speed services are significantly higher than the European average. High prices for volume and high-speed services are particularly problematic because, according to 2022 data, mobile internet consumption is growing dramatically, reaching an average of 12GB.
“As long as high prices for high-speed and volumetric internet remain high in Georgia, consumers will continue to be forced to choose low-speed and low-volume services or pay significantly higher, premium prices,” concluded the Communications Commission.
Low-volume mobile service packages for individual subscribers in Georgia, such as 5GB and unlimited minutes, are 1% more expensive than in Europe, and the price of 5GB internet and 577 minutes is 23% higher. One of the most consumed optical internet packages for individual subscribers, 25Mbps in Tbilisi, is 5% cheaper than the average price of similar services in European countries, and 17% cheaper in the regions.
The comparative analysis of the combined offers showed that for more than half a million families living in Tbilisi and the regions, the cost of the cheapest Internet and TV packages is lower than the European average.
In hindsight, the company’s announcements at MWC signalled that the pivot to AI-over-all was already irreversible
While Microsoft’s decision to cut hundreds of jobs at its Azure cloud unit was reported widely as a general tech malaise, the focus areas of the cuts signalled a pivot in how it was approaching the telco market. According to the reports, first raised by Business Insider, Microsoft’s Azure for Operators layoffs would reportedly involve up to 1,500 jobs.
The other impact area was the mixed reality Mission Engineering team which is not too much of a surprise given the real world has triumphed over the metaverse, at least in enterprise terms. Korea Telecom is only the latest to quietly withdraw from enterprise metaverse.
With Azure for Operators, Microsoft wanted to build an ecosystem of partners that could run AI-driven networks on a carrier-grade, hybrid cloud platform. But carrier networks are littered with legacy tech so the efforts to integrate multiple, diverse partners to apply multiple diverse back-office solutions to multi-diverse operators would always live and die by its unit sales.
Microsoft’s acquisition of Affirmed, plus AT&T moving its 5G core network into Azure, were clear milestones the cloud provider understood how to shift packet core workloads to the cloud. But while Affirmed gave it the ability to extend cloud-based software defined networking into the world of 5G connectivity, Ericsson, Nokia, Samsung and the Open RAN proponents were on the similar paths and have the incumbency.
AI-over-all
In a follow up story, Business Insider [subscription required] confirmed that Microsoft was pivoting to AI for pretty much everything – Microsoft has announced billions in data centre/AI investments globally in the past couple of months alone. Business Insider reported a leaked memo from exec Jason Zander suggesting the layoffs were a direct result of its AI investment.
The memo reportedly also confirmed Microsoft would halt previews for Azure Operator 5G Core (AO5GC) and Azure Operator Call Protection. In addition, some remaining team members in Azure Operator Nexus are heading off to Microsoft’s Azure Edge and platform product line teams. Given the fact that Azure Operator Nexus had been described by some in the industry as the new operating system for telcos, it is easy to discern the complete change in thinking around what telco OSS/BSS/orchestration and so on is going to look like.
MWC hints
The public preview of Azure Operator Call Protection, the service that uses AI to help protect consumers from scam calls, was literally only previewed at Mobile World Congress. At the time, Microsoft announced BT was trialling the service which used real-time analysis of voice content, alerting consumers who opt into the service when there is suspicious in-call activity.
E& UAE was named as a customer of Azure Operator Nexus hybrid cloud platform that enabled operators to run workloads on-premises or on Azure. AT&T was also cited although the latter moved its 5G core to Microsoft’s Cloud. Three UK was mentioned as a user of Azure Operator Insights.
Microsoft also launched Azure Programmable Connectivity (APC), which provides a unified, standard interface across operators’ networks and seamless access to Open Gateway for developers to create cloud and edge-native applications that interact with the intelligence of the network. This will presumably be wound up.
One service that will probably survive in a different form – only had a limited preview at MWC – was Microsoft’s Copilot in Azure Operator Insights which allows engineers to use Copilot to interact with network insights using natural language and receive simple explanations of what the data means and possible actions to take. That said, this is pretty similar territory to the Global Telco AI Alliance.
Too niche?
The point is that many operators are deeply embedded with Microsoft already – they are past the point of accepting that their compute workloads need to run in the cloud. And its Microsoft’s OpenAI work that operators can not only embed in their back-office processes but can potentially monetise as well. In Match, for example, Deutsche Telekom signed up its first Business GPT customer.
Telefónica evolved its digital framework for the creation of advanced services, Kernel, by adding new Microsoft Azure AI services to power every relevant use case with generative AI capabilities at scale. Meanwhile, Vodafone said it will apply Microsoft Azure OpenAI to deliver frictionless, real-time, proactive and “hyper-personalised” experiences across all Vodafone customer touchpoints, including its digital assistant TOBi (available in 13 countries).
Operators are going to continue to embed AI wherever they can get a cost benefit, regardless of whether Azure for Operators exists or not. Microsoft’s decision to cull in this area just reminds operators that sorting out their back-office IT stack is not straightforward and certainly not yet standardised. And it does not diminish the impact Microsoft will have with AI in telco environments.
Partner content: Why does personalisation matter so much?
While personalisation is not an entirely new topic, there is currently massive hype around it. This is due to service providers having realised that they need to be more cost-effective in engaging with their customers and because customers now demand to be treated individually and in a personalised way.
McKinsey & Company[1] states that 71% of consumers expect personalisation, while 76% of those same consumers get frustrated if they cannot find personalisation in the brands they interact with, leading to churn. The study also claims that 75% of consumers switched to a different brand because they weren’t satisfied with the experience they were getting. This is even more critical in the aggressive market service providers face today.
The study concludes that personalisation pays off, as brands that capture more value from personalisation will grow faster and can earn 40% more revenue from personalised marketing actions or tactics.
Although some service providers are already delivering hyper-personalisation use cases to their customers, most businesses are still experiencing segmented personalisation. Nevertheless, regardless of the adopted approach, it is easy to understand that, as service providers become more mature in personalisation, it is possible to improve the response rates, increase the conversion rates and lift the customer lifetime value. But one question remains: are they taking action to harvest all the available benefits?
Yes, but not enough!
Several market trends have been adopted in recent years, bringing service providers to the level where they currently stand:
Customers are segmented into groups by their similarities and offered similar value propositions and experiences. However, what is suitable for the group might not be good for the individual.
Customers’ deflection to non-assisted channels brings efficiency and allows them to reduce operational costs. Nonetheless, all the interactions are detached, impersonal, and less human.
Service providers are automating sequential journeys, ignoring the fact that sequential journeys are dead! The new way customers interact with their brands is through a set of non-linear moments and different touchpoints that need to be correlated and orchestrated.
The previous trends already brought benefits and were applicable in a particular context. Today, new approaches are required for service providers to progress, improve, and move to the next level.
Market challenges Achieving a hyper degree of personalisation can be challenging – it requires investment, time, and the right strategy. Furthermore, it is not a straightforward path, and it’s essential to understand the business and technical challenges that service providers must address and overcome to succeed. Challenges for which Hyper-Personalisation is the answer.
Service providers face intense pressure to deliver superior customer experiences to differentiate from competitors. With products, services, and prices being similar across the board, customer experience becomes the key differentiator. Lack of differentiation leads to increased competition and pressure to retain customers, making customer satisfaction, loyalty, and trust critical. Additionally, service providers need to boost annual revenue in a saturated market, which demands cost-effective marketing strategies. Mass marketing investments often yield low returns, highlighting the need for more precise and efficient customer engagement.
Large customer bases
Although the telecommunications industry has the advantage of having large customer bases and data, the challenge lies not in data availability but in extracting actionable insights from it. Telcos face the technical challenge of creating a unified customer profile from disparate information sources. This requires integrating online and offline data to achieve a real-time, comprehensive view of customer knowledge while minimising reliance on third-party data.
A critical aspect of this challenge is the ability to correlate digital behaviours with customer intent. By anticipating and predicting customer intentions, service providers can proactively engage with customers, meeting their expectations with differentiated services. Furthermore, transforming this data and intent into relevant, actionable insights is essential for delivering personalised services. These insights must be generated in real time due to the rapidly changing nature of customer behaviour, ensuring that the information remains accurate and useful.
The hyper-personalisation approach
Figure 1 – Hyper-personalisation platforms
The hyper-personalisation approach must bridge Digital with Data and Cognitive capabilities to deliver unique, in-context and highly relevant experiences to individuals.
This approach requires three main platforms as the foundations of the solution:
Customer Data Platform is the backbone of the customer profile. Customer-related data collect different and relevant real-world events.
Data Insights Platform represents the decision engines that run on top of the customer data to provide relevant and actionable insights, with two primary responsibilities — campaign management and AI & ML.
Digital Experience Platform, used to activate and deliver the experience in the digital channels as per the insights from the Data Insights Platform.
Most of the industry is still focused on segmented personalisation, delivering experiences and recommendations to grouped users considering predefined rules. This can be seen as the starting point or the primary solution, but it depends on the maturity level of each service provider. Usually, it is the domain in which most of them are working.
With hyper-personalisation, service providers can scale up the personalisation leveraged by AI & ML to deliver experiences and recommendations in real time to individuals, creating a one-to-one personalisation.
To achieve this, these one-to-one engagements must be tailored to specific micro-moments. One of the most used frameworks is that of Google[2], which identifies three main micro-moments:
“I want to know” – Related to the market. When the customer wants to find something.
“I want to buy” – Related to the selling moment. When the customer wants to buy something.
“I want to do” – Related to engaging or with the care moment. When the customer needs to do something or solve a specific issue.
For each of these specific micro-moments, service providers can instantiate a hyper-personalisation solution that addresses the particularities of each moment with a greater focus on the customer’s specific intent:
Smart Content Experiences: Delivers tailored and behaviour-driven personalisation leveraged on continuous data collection to provide unique content according to customer intent in real time, leveraged by the Digital Experience Platform.
One-to-one Recommendations: Delivers contextualised one-to-one products and offers recommendations based on customers’ real-time interaction history & preferences, leveraged by Campaign Management or AI & ML models.
Cognitive Care: Uses AI & ML to predict customers’ intentions and recommend the best experience to engage with them proactively.
The benefits
One of the main benefits of using highly personalised experiences is that service providers can maximise the value of each customer throughout their lifecycle.
Figure 2 – Improving customer lifetime value through hyper-personalisation
This improvement isn’t achieved by engaging with customers only at the beginning and end of their lifecycle, as does the classic CVM approach, but constantly throughout it all. Following this strategy maximises the Customer Lifetime Value and optimises internal costs, especially acquisition and retention costs.
Additionally, according to McKinsey & Company[3], hyper-personalisation instantiation can result in clear and tangible business benefits:
Engaging directly with customers and avoiding mass campaigns can reduce up to 50% of acquisition costs. This, paired with direct savings from the reduction in marketing investments, enables 10–30% better cost-effectiveness.
The buying process becomes more natural when directly targeting customers’ emotions and addressing their preferences, interests, and tastes. This approach allows up to 2.5x better conversion rates. Besides this, continuous and closed engagement during the care period, even if representing a slight increase in investment, delivering contextualised and relevant up-sell and cross-sell recommendations, contributes 5–15% uplift revenues.
Proactively engaging with customers during the care period enables predicting and preventing possible customer service issues. This type of engagement allows up to 30% improvement in customer retention. This proactive care approach can increase the Net Promoter Score by up to 20%, enabling a better overall customer experience.
Article originally published by Henry Stewart Publications in the Journal of Digital Banking, Volume 8, Number 3.
About the author (pictured above)
Luís Coelho is Head of Hyper-Personalisation Offer at Celfocus and a Senior Manager at Celfocus Offer Development & Innovation team. He has a long experience in IT projects’ delivery and pre-sales activities for the Telco & Financial Services industries, gathered from more than 20 years’ real-life and hands-on practice delivering mission-critical solutions.
Currently, Luís is the Head of Hyper-Personalisation Offer with a focus of providing business and technical solutions to extract value from data for successful digital enablement. This is achieved by bridging digital, data and cognitive capabilities to deliver unique and highly relevant experiences to individuals, allowing service providers to reach customers more efficiently.
Meanwhile, the German government offloads €2.5bn of DT shares (a stake of about 2.2%) to raise cash for the country’s neglected railway
Deutsche Telekom (DT) has installed a private 5G Standalone (5G SA) campus network for RTL Deutschland, just in time for the European Football Championship. The entire infrastructure, from the antennas and active system technology to the core network, comes from Ericsson and is installed locally on the site in Cologne-Deutz.
The 5G campus network is installed in two RTL studios and the so-called mall, plus RTL’s visitors’ car park in Cologne-Deutz has 5G. Altogether the network covers an area of more than 35,000 square meters for uninterrupted coverage indoors and outdoors. DT says the private network will ensure “a flawless live broadcast”, even if the mobile cells of the public network are busy.
DT and RTL have already successfully tested live broadcast via the public Telekom 5G SA network with network slicing for reporting on the move.
All data traffic remains in the local campus network. A total of six antennas were installed for the campus network at RTL Deutschland: two outdoors, one of which is used on a temporary basis, and four more indoors.
The upload capacity is about 500Mbps are for live productions – the delay rate must not exceed 25 milliseconds. The 5G SA network operates on frequencies reserved for RTL Deutschland in the 3.7 to 3.8 GHz range; up to 100MHz bandwidth is exclusively available to the broadcaster.
RTL Deutschland can adapt the private network and manage various functions on-demand. For example, specific applications’ data traffic can be prioritised as required. The local core network has a redundant architecture ensuring the network continues reliably even in the event of an interruption to the cloud-based management portal.
Selling DT to fund rail
Germany’s government has sold a €2.5 billion stake in DT. The country’s media reports this the latest such move by the government to raise cash from selling off corporate holdings.
The German state-owned bank KfW offered institutional investors 110 million shares on Monday. The combined holding of KfW and the government is now 27.8%. They remain DT’s largest shareholders.
DT responded saying it will increase its share buyback programme by €600 million. The programme was launched in November as a means of paying back shareholders after DT raised funds to increase its stake in its US subsidiary, T-Mobile.
Need to raise funds
The Finance Ministry said the net proceeds will be used to increase the government’s equity of state rail operator, Deutsche Bahn, and expand the country’s rail infrastructure. Deutsche Bahn itself Deutsche Bahn is looking for bids for its DB Schenker logistics unit, which could bring in more than €15 billion.
“Due to the receptive stock market environment, the placement was successfully completed,” the Ministry said in a statement. top court last year.
According to The Business Times, “Berlin is trying to come up with savings measures without compromising efforts to lift Europe’s biggest economy out of a slump or modernise and expand the armed forces after years of neglect.”
The pilot proved uplink coverage on the higher frequency is compatible with existing inter-site distances
Nokia and Telia have successfully completed a field pilot in the upper 6GHz spectrum range. They say it will “add crucial capacity and coverage to existing macro cell sites in dense urban environments for next-generation 5G-Advanced and 6G networks”.
Due to rapidly rising volumes of data traffic, most operators will need to increase their TDD mid-band spectrum holdings in the second half of the decade. The allocation of the upper 6GHz spectrum for mobile services – which could mean 200MHz of mid-band TDD spectrum per operator – was agreed at the World Radio Conference 2023. It is especially important in markets where the 7.2-8.4 GHz range will not be available.
Pilot explores
The pilot explored whether the uplink coverage on the new, higher frequency is compatible with existing inter-site distances. The companies tested the upper part of the band (n104) and used a 3.5GHz massive MIMO cell of the same radio frequencies across various distances to replicate different real-world scenarios.
Field tests confirmed the macro-grid of the upper 6GHz spectrum can be used with Massive MIMO. It showed that massive capacity can be added in urban areas, where there is higher demand for TDD broadband, and high throughput can be achieved in suburban or rural areas. This offers operators an evolution path to 5G-Advanced and 6G.
The trial used a Massive MIMO antenna based on Nokia’s AirScale Habrok radio with a test terminal from MediaTek with integrated antennas.
Additional spectrum
Luciana Camargos, Head of Spectrum at the GSMA, said, “Mobile operators need additional spectrum to serve growing connectivity needs, so commercialisation of the upper 6 GHz band will be an important tool in enabling the digital ambitions of each country.”
Stefan Jäverbring, Group CTO at Telia, added,“As our customers generate increasing amounts of mobile traffic, it is essential to have further access to mid-band TDD spectrum to enhance digitalization in our markets and serve our customers appropriately.
“This field test with our partner, Nokia, is an important step in demonstrating how this can be done sustainably, as it would be possible to use our existing site grid. In this way, deployment would be faster and have less environmental impact, creating fewer carbon emissions than the alternative of adding capacity by building more new sites.”
“This field pilot demonstrates that Nokia is ready to help mobile operators integrate the bandwidth of these new spectrum allocations seamlessly into their existing network allowing them to provide coverage from the existing macro cell sites on the higher frequency bands,” Mark Atkinson, Head of RAN at Nokia, concluded.
The cloudco intends to provide AI competence training to 2.4% of the population, reflecting the importance of AI in Sweden, and boost green credentials
Microsoft is to invest SEK33.7 billion (€2.962 billion) in Sweden over the next two years on cloud infrastructure and the development of AI.
This is the biggest investment yet by Microsoft in the Nordics and will expand Microsoft’s footprint, deploying “20,000 of the most advanced GPUs to its existing data center regions in Sandviken, Gävle and Staffanstorp,” according to this post, in Swedish.
Sweden is a key market
The post cited the AI Index report from Stanford University that shows Sweden has the third largest requirement globally for AI skills (see graph below). Microsoft intends to address this by raising competence, security and knowledge regarding AI within organisations, schools, universities, the public sector and society at large. It plans to provide competence training for 250,000 people over three years, which corresponds to about 2.4% of the population.
Source: Artificial Intelligence Index Report 2024, published by Stanford University
The post says, “To guide the skills programs, Microsoft is establishing an AI Insights Council that brings together leaders in academia, business and the public sector. The competence program will be run in close collaboration with partners and representatives from, among others, the Employment Service, Linköping University, TechSverige and partners within the Swedish Microsoft’s Responsible AI Innovation Center.”
Green agenda
Microsoft noted that working with the state-owned power company, Vattenfall, it can operate its datacentre regions on 100% fossil-free energy and water is only for humidity, not cooling, and most of that is collected rain water.
The software giant also points out it has completed a pilot with Saft, a subsidiary of TotalEnergies, using battery energy storage systems (BESS) for back-up instead of diesel generators. Apparently this is an important milestone towards Microsoft’s diesel-free data centres by 2030.
Brad Smith, Vice Chair and President at Microsoft, was also quoted saying that the company is “committed to increasing accessibility and accelerating the introduction of AI throughout the Nordics, including Denmark, Finland, Iceland, Norway and Sweden. We will connect our AI investments to programs and partnerships that support individuals, businesses and long-term economic growth across the region.”
Operator said it now has more than AI 400 professionals located across ten specialised centres
Telefónica’s digital business unit Telefónica Tech said it now has more than 400 professionals with AI expertise distributed in Spain, the UK, Central Europe, Brazil and Hispam (Latam) to provide customers with a service tailored to local needs and consolidate its position as a benchmark technology partner
These professionals are currently distributed in the 10 specialised centres that the company has created in Spain (Madrid, Barcelona, Valencia and Valladolid), the United Kingdom (London), Central Europe (Slovenia and Austria), Brazil (Sao Paulo) and the Hispam region (Santiago de Chile and Mexico City) to provide a full service and extended hours to customers.
“We have numerous technological solutions to optimise the processes of organisations based on artificial intelligence and the enormous potential it offers when combined with other technologies such as the IoT, blockchain and cybersecurity,” said Telefónica Tech director of the artificial intelligence and data business unit Elena Gil Lizasoain.
“The global knowledge and experience of our teams is allowing us, for example, to help companies predict the demand for their products, automate processes, improve their decision making and personalise their customers’ experience,” she added.
The operator has also signed deals with hyperscalers that include AI elements. For example, last February, Microsoft announced its Responsible AI Innovation Center (RAIIC) in Spain, in which Telefónica Tech collaborates, along with 15 other technology partners, to promote the adoption and secure use of artificial intelligence. And two weeks ago, Telefónica Tech announced a new agreement with Microsoft to bring next-generation cybersecurity services, including automation and AI elements, to enterprises around the world.
Last week, the operator renewed and extending its partnership with Google Cloud. Telefónica is already using Google Cloud internally to advance in its own digital transformation, including using Google Cloud’s technologies in areas such as IT through its “Go to Cloud” programme. The unit also has engaged AWS for wider cloud initiatives – nearly 80% of its workforce earned over 400 AWS Certifications and more than 800 AWS Partner Accreditations. Telefónica Tech also launched an AWS Cloud Camp, known as Escuela 42, to create new pools of cloud talent in Spain.
Sector targets
Telefónica Tech has artificial intelligence products and services adapted to the specific needs of different sectors of activity, with public administration, health, industry and sport being the areas in which some of the most relevant projects have recently been developed.
The operator said it is collaborating on projects with several employment departments to help improve the functioning of the labour market with a model based on professional skills, which allows predicting those that will be most in demand and guiding the unemployed with training recommendations.
In the healthcare sector, Telefónica Tech has developed a joint product with the multinational Solventum (formerly 3M Healthcare) to boost the optimisation of hospital emergency services. AI makes it possible to predict the number of patients who will attend in the coming days, what specialties they will require and how many of them may need to be admitted. In industry, Telefónica Tech is applying AI to exploit the data generated in the factory to facilitate demand planning, predictive maintenance and quality control, avoiding production stoppages and minimising the consumption of raw materials.
Sport is another of the areas in which Telefónica Tech is intensively applying this technology together with IoT to help technical teams to improve athletes’ training, prevent possible injuries and prepare strategies for each competition. For example, the company is helping the Royal Spanish Athletics Federation to improve the performance of athletes using inertial measurement devices and the application of AI algorithms and advanced analytics on the data collected.
The rural sector wants to embrace digitisation but internet connectivity is seen as one of the strongest obstacles to doing so
Almost half of the farms in Germany (47%) are currently dealing with possible applications of AI, while one in ten companies (9%) is already using it, and a further 38% are planning or discussing it. However, in a study carried out by Germany’s digital association Bitkom and German Agricultural Society (DLG) looking at the digitisation of agriculture, more than half of respondents (51%) consider inadequate Internet coverage to be one of the strongest obstacles to achieving their AI aspirations.
This snapshot is important given Germany is among the top four agricultural producers in the European Union – producing €50 billion worth of agricultural products each year. Germany is also the second largest organic market in the world (behind the United States) and the survey showed 4 out of 5 companies concluding that digitisation enables more environmentally friendly production.
Europe’s governments understand that connecting the rural sector is sometimes uneconomic, but many will need to examine the opportunity cost of not supporting connectivity incentives to bring this about and delivering the connected farm.
Despite more than half of the respondents complaining about a lack of involvement in the planning of political measures, the opportunities of digitiation are being seized. Whether sensor technology, robotics or digital field index, the use of digital technologies has generally increased in the past two years: GPS-controlled agricultural machinery is the most common, which is already used by 69%. Two years ago, it was only 58%. This is followed by digital field files or cow or sow planners with 68%. This is used to track breeding cycles in animal husbandry, among other things. Farm or herd management systems are now used by 46%. of farms, compared to only 32% in 2022.
An opportunity for the sector
In general, a large majority (79%) of farmers see digitization as an opportunity for their farm. Only 15% see it as a risk, and for 6% digitization has no influence on operations. The biggest advantages that farmers personally perceive on their farms through digital applications are time savings (69%), higher efficiency in production (61%) followed by physical relief (57%).
Specifically, 91% of farmers believe that digital technologies help to save fertilizers, pesticides and other resources. 69% say they can contribute to increasing animal welfare. 67% say that with the help of digital technologies, farms can reduce costs in the long term and 60% see this as an improvement in the quality of agricultural products.
Sensors, drones and management systems
36% of farms already rely on applications for the site-specific application of fertilisers (2022: 30%) and 30% for the application of plant protection products (2022: 23%). Sensor technology in animal husbandry and crop production is used by 28% (2022: 22%). Predictive maintenance, for example for agricultural machinery, is used by a quarter (25%) (2022: 19%). 24% use automatic feeders or intelligent feeding systems (2022: 24%). Drones are also used by just under a quarter (23%), compared to 19% in 2022. 12% are already relying on robotics (2022: 10%). A total of 90% of the companies use at least one of these digital solutions.
Challenges to the digital farm
While politicians can help on the investment front, farmers want the industry to reduce the complexity of the tech. More than half (54%) of respondents said digitisation is a challenge, despite the potential rewards.
When asked about the strongest obstacles to the digitisation of agriculture, most (75%) cite high investment costs. This is followed by concerns about more bureaucracy at 61% and insufficiently standardised interfaces and networking of systems at 59%. Half (52 %) of the companies complain about a lack of involvement in the planning of political measures. 51% consider inadequate Internet coverage to be one of the strongest obstacles.
This is followed by concerns about a loss of data sovereignty or a high level of complexity of digital systems with 49% each. 47% are concerned about IT security and 41% see a lack of digital skills as an obstacle.
Where AI will impact
The larger the farm, the more intensive the use and engagement with AI: while only 27% of farms with 20 to 49 ha use, plan or discuss the use of AI, the figure is 38% among farms with 50 to 99 ha and even 52% for large farms of 99 ha or more. “Agriculture is one of the pioneers of AI and is ahead of most other industries. AI can massively relieve the burden on farms, leaving farmers more time for other tasks. Smaller companies in particular should make greater use of the possibilities of AI,” said Bitkom CEO Dr Bernhard Rohleder.
DLG VP Prof. Dr Till Meinel, Institute of Construction and Agricultural Machinery Technology, Cologne (IBL) added: “The use of AI is not a trend, but increasingly an absolute necessity due to the diverse burdens on farm managers.”
The greatest potential for AI use in agriculture is seen in forecasts and crop protection, but also in office work: 54% of farms that are already using AI, planning or discussing it do so for climate and weather forecasts, 36% for market analyses or price predictions, 28% each for harvest and production planning or yield forecasts. 46% of farms that use, plan or discuss AI want to improve crop protection, for example through disease diagnoses, and 20% want to improve health monitoring in livestock farming.
But AI is also being planned, discussed or already used away from the barn and field, at 4 out of 10 companies (39%) for everyday office work such as administrative activities.
